# FluxML/Flux.jl

Relax! Flux is the ML library that doesn't make you tensor

Repository: https://github.com/FluxML/Flux.jl
Canonical: https://ross.abutalabs.com/products/fluxjl
Homepage: https://fluxml.ai/
Language: Julia
License: NOASSERTION
License Family: other
Topics: flux, machine-learning, neural-networks, deep-learning, data-science, the-human-brain
Last push: 2026-08-16T08:15:53+00:00

## Health v2 (maintenance only)
Score: 98/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 96, longevity 100
- inputs: {"age_days": 3806, "days_push": 17, "days_rel": 27, "gap_med": 15.0, "n_releases_24m": 21}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4739, forks 624 (observed 2026-08-28T04:08:57.785916+00:00)

## What it is
Flux.jl is a machine learning library written entirely in Julia, providing lightweight abstractions over Julia's native GPU support and automatic differentiation. It lets any parameterized Julia function serve as a model while remaining fully hackable, with an ecosystem covering computer vision, NLP, graph neural networks, and SciML.

## Use cases
- train neural networks in Julia
- define models as plain Julia functions with automatic differentiation
- train models on GPU with CUDA.jl
- build neural ODEs and scientific ML models
- use pretrained computer vision or transformer models
- fit a function to data with gradient descent

## When to choose
- you work in the Julia ecosystem and want a pure-Julia ML stack
- you need hackable, differentiable models mixing neural nets with scientific computing
- you want GPU-accelerated training without leaving Julia

## When to avoid
- you need the mature tooling and ecosystem of PyTorch or TensorFlow
- your team works primarily in Python
- you need extensive production deployment tooling

## Facets
- artifact type: library
- maturity: stable
- function: machine-learning, deep-learning, llm-training
- domain: machine-learning, deep-learning, data-science
- platform: cross-platform
- tags: neural-networks, differentiable-programming, pure-julia, automatic-differentiation, model-zoo, julia, gpu

## Member repositories
- FluxML/Flux.jl (main) score 98

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:57.785916+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:18:58.293933+00:00, confidence not recorded.
  - readme: https://github.com/FluxML/Flux.jl (fetched 2026-08-28T04:08:57.785916+00:00, sha 13ba82252db9)
  - homepage: https://fluxml.ai/ (fetched 2026-08-29T09:02:29.476947+00:00, sha c95d91c263ec)
  - site_page: https://fluxml.ai/Flux.jl (fetched 2026-08-29T09:02:29.479548+00:00, sha ac03d6c9419b)
- Data as of 2026-08-30T08:39:29.467469+00:00.